Project Summary Dan P. Ly, MD, MPP, is an inpatient teaching attending physician in the Veterans Affairs (VA) Boston Healthcare System and a PhD candidate in health policy (economics track) at Harvard University whose long- term interest is in improving the quality of care for vulnerable populations. This postdoctoral fellowship will help Dr. Ly develop into an independent investigator with expertise in physician decision-making. Dr. Ly's extensive programming experience with large databases and the substantial formal training he has received have prepared him to perform the proposed research. This proposal, on how cognitive biases among physicians may be exacerbated when caring for older adults with Alzheimer's Disease and Related Dementias (ADRD) and for minority patients, builds on Dr. Ly's prior work on health care disparities by testing for a mechanism for such disparities. The training plan will build upon prior training through three goals: 1) gain guided experience with the care of older adults from specialists in dementia and geriatrics; 2) develop expertise in helping doctors make better decisions; and 3) acquire skills in methodology (e.g., qualitative and experimental methods) complementary to Dr. Ly's training in large database analysis. Dr. Ly will be mentored by Dr. David Cutler, a nationally-renowned health economist who is an eager and experienced mentor. His co-mentor is Dr. Anupam Jena, a physician and health economist with physician behavior expertise. Dr. Ly has assembled an advisory committee with expertises in gerontology, decision-making, statistics, management, and informatics. Quality of care is known to be suboptimal, particularly for older adults. As the population ages, it is important to identify barriers to delivering high-quality care to older adults. One set of barriers may be physician cognitive biases that may occur more often when caring for such patients, especially patients with conditions like ADRD that may make communication difficult. Some cognitive biases may also worsen when caring for racial and ethnic minority patients. Little is known about consequences of cognitive biases on a national scale for these vulnerable populations because the factors physicians weigh are often not in the claims data commonly used. Dr. Ly will use rich, national electronic health record data from the VA and focus on a common scenario: assessing patients with shortness of breath for the risk of pulmonary embolism (PE). His project has three aims. First, he will examine whether physicians pay attention to objective clinical factors validated to correlate with PE risk and whether they also focus on arguably extraneous factors, such as diagnoses of recent patients seen (Aim 1). Second, he will determine whether physician attention on these objective factors is significantly lower for patients with ADRD (Aim 2). Third, he will examine whether this attention is also significantly lower for minority patients (Aim 3). This research may help explain deficiencies in quality of care for vulnerable populations and inform the design of future interventions to counter cognitive biases and improve care.